Many strategies for gene therapy of solid tumors involve the use of replication-defective Moloney murine leukemia (MLV)-based retroviral vectors, but efficacy has been limited due to lack of adequate tumor transduction. Gene transfer using replication-competent retroviral vectors would be more efficient, as each transduced tumor cell produces more vectors capable of initiating further infection events. We have devised novel retroviral replicating vectors (RRV) capable of highly efficient gene delivery to solid tumor in vivo. We have demonstrated that vector spread is restricted to the tumor itself due to restriction by innate immunity in normal cells, and the intrinsic inability of the virus to infect ost-mitotic normal cells. We have enrolled 17 subjects with recurrent high grade glioma (HGG) in multi-center Phase I trials (www.clinicaltrials.gov:NCT01156584, NCT01470794). The RRV in these studies, Toca 511 (vocimagene amiretrorepvec, or RRV-CD in this proposal), expresses an optimized yeast cytosine deaminase (CD) which converts the prodrug 5-fluorocytosine (5-FC) into the classic chemotherapy drug 5-fluorouracil (5- FU). 5-FU is often used in chemotherapy combinations, but agents that directly inhibit cell proliferation interfere with vecto spread. We propose to test combinations of different prodrug activator RRV to inhibit tumor without affecting vector spread. This is made feasible by our discovery that RRV infected tumor cells can be reinfected in vivo. We have now constructed RRV expressing an optimized Herpes simplex thymidine kinase (HSV-TK) prodrug activator gene (RRV-TK in this proposal). We propose to perform validation studies with the newly developed RRV-TK vector both in vitro and in glioma models in vivo. We will evaluate the potential for dual RRV gene therapy combining RRV-CD and RRV-TK. We will 1) test the transduction efficiency, transgene expression level, replicative stability, and cytotoxic potency of RRV-TK, in vitro and in vivo, and 2) determine the optimal parameters for combination therapy, using both RRV-CD and RRV-TK, in vitro and in glioma models in vivo. The information from these studies will be used to develop future clinical trial protocols and, like combination chemotherapy, offers the potential for synergistic therapeutic benefit and more efficient tumor eradication than single-agent therapy with either vector alone. Furthermore, as the infected tumor cells themselves produce their own chemotherapeutic drugs, the adverse effects of systemic chemotherapy are avoided. PUBLIC HEALTH RELEVANCE: Many researchers have used 'crippled' viruses (termed 'vectors') that can infect and modify tumor cells as cancer therapies. However such therapies have been shown to have only limited beneficial effects, because too many tumor cells never get infected. We have devised a new approach, using replicating viruses to achieve highly efficient gene transfer to tumor cells in a highly selective manner (i.e., the virus is selective fr cancer cells and will not infect normal cells). The infected tumor cell becomes a virus-producing cell, sustaining further infection. Our first vector is being tested in a multi-center clinical tril in patients with brain tumor. This proposal seeks to develop a second version of this type of vector to be used in combination with the first. These two gene therapy vectors cause the cancer cells to produce their own chemotherapy drugs and the second vector potentially enables non-invasive imaging of vector infection of brain tumors. If successful in these preclincial studies, this approach can be translated into the clinic and has the potential to reduce suffering and death in brain cancer patients. It may also be applicable to brain metastases caused by other malignancies.